Exploiting the Therapeutic Potential of Plants

Technology Networks recently interviewed Viswa Colluru, founder and CEO of Enveda Biosciences, to find out how the company is “harnessing the complexity of the natural world to tackle today’s biggest healthcare challenges”. Colluru discusses the therapeutic potential of plant-derived molecules for drug discovery and elaborates on how the company’s platform is helping to drive the next-generation of small molecule therapeutics.

Laura Lansdowne (LL): Tell me a little bit about Enveda and how the company was founded?

Viswa Colluru (VC): I founded Enveda Biosciences in 2019 with Pieter Dorrestein, Ph.D., a world-leading metabolomics researcher. My educational background in cellular and molecular biology and later my professional background with automated biology and computational methods to accelerate drug discovery at Recursion really primed me to think about the ways in which we could approach drug discovery in completely new ways using technology.

I realized that the fundamental problem in drug discovery was that a lot of the science that worked in the laboratory did not work in a clinical trial setting. If you think about it, this isn’t surprising given the complexity of the real world and controlled conditions in the laboratory, and the differences between mice and humans. Throughout my life, I had personal experiences through myself and family members where I saw the benefits of medicinal plants and how they were used in India, where I grew up. In fact, drugs that have been used to treat billions of patients, like artemisinin (Nobel Prize winning discovery for the treatment of malaria), aspirin (needs no introduction!), warfarin, morphine, metformin (used by hundreds of millions of diabetics), and cannabidiol (recently approved by the US Food and Drug Administration (FDA) for treatment of rare forms of epilepsy), were all discovered based on “traditional uses” of medicinal plants. The exciting part was that there were thousands of plants that were untapped for modern drug discovery, largely due to the unique technical challenges of analyzing these plants. When I read Pieter’s paper on molecular networking, it really got me thinking about how this and other tools could be used to improve drug discovery. All these aspects of my background came together and allowed me to think about how we can unlock nature’s secrets to deliver the next 100 aspirin!

The Enveda platform is the first of its kind, systematically enabling the translation of molecules found in medicinal plants into new drugs for challenging diseases. Our platform harnesses nature's complexity with the help of cutting-edge advancements in knowledge graphs, machine learning (ML) and metabolomics. We recently raised our seed round of $4.9 million and are now working on growing our platform and advancing our lead candidate through preclinical development.

LL: How does your platform work?

VC: We are building the largest knowledge graph of medicinal plants in the world. Knowledge graphs are a special kind of powerful database, uses include Google's method to answer questions and Netflix to power your recommendations. This includes anthropological, biology and chemical data that is pulled from recorded knowledge from human use and new surveys, existing literature and the most comprehensive set of in-house multi-omics and metabolomics studies on these plants. We use ML to study all of this data and propose high-potential hypotheses around new chemical scaffolds that might have therapeutic potential. We then test these molecules in vitro and in vivo and feed the data from these studies back into the database, making our platform smarter with every experiment. Our promising candidates are then further improved on and made into potent and patentable clinical candidates.

LL: Why are medicinal plants such an asset in drug discovery?

VC: There are three main reasons.

1. Plants have proven to be a highly reliable source of new medicines. For nearly 50,000 years, we have been harnessing the therapeutic potential of plants across cultures. Plants gave us some of the first modern medicines. Aspirin, quinine, metformin, artemisinin and morphine are all inspired by molecules found in plants that were used as medicine. In fact, from 1981 to 2020, about 33 percent of regulatory approved small molecule drugs were based on molecules found in the natural world.

2. Plants represent massive, and highly relevant chemical diversity. There are about 400,000 plants known today, collectively containing millions of molecules. Plants are also chemists beyond comparison, producing secondary metabolites that have evolved under intense evolutionary pressure to interact with conserved biological domains across microbes and herbivores, including mammals.

3. In spite of the validated promise, plants remain almost entirely untapped. Only ~ 5% of plants are estimated to have been studied for potential therapeutics. Moreover, ~ 99% of the molecules found in even well-studied plants are uncharacterized. Until recently, we really didn’t have the tools and technologies available to us to better understand the chemistry of these plants. Because of this, there is tremendous potential that our technologies can capitalize on.

LL: How are Enveda’s technologies helping researchers to realize the potential of plant chemistry?

VC: We combine zero to one type of advancements across multiple fields. This includes tools such as molecular networking (used to visualize tandem mass spectrometry data) and metabolomics (the study of metabolites produced by an organisms’ chemical processes). We have also made major advancements in areas such as ML, the computational encoding of biological knowledge and knowledge graphs. These tools and advancements make identifying the mechanism of action and active molecules easier and also allow us to visualize and synthesize all the available data more efficiently.

But even as these tools have become more widely available and more readily adopted, without a systematized body of knowledge, like Enveda’s database, it is hard to even comprehend where to begin. While we may see that a plant can decrease a fever, we have no clue if that means it is treating a parasitic, bacterial or viral infection, or because it stimulates the host immune system, because we don’t have the anthropological, biological and chemical data available to us in one place – or at all.

Now that we do have the tools and the data available in one place, the challenge is to convince the industry that now is the right time to revisit this type of drug discovery.

LL: Why will that be a challenge?

VC: Primarily because our approach strays from the status quo. Existing drug discovery hit identification workflows focus on synthetic libraries, which were created after the molecular biology revolution. Since this revolution, the industry has adopted a target-based approach to drug discovery, where researchers focus on first identifying a protein responsible for a disease, and then finding a molecule that can modulate that protein target. Thus, screening libraries are loaded with compounds prepared for a finite number of specific molecular targets. While successful, this screening approach explores a similar finite, and often redundant, chemical space. While natural products significantly expand that chemical diversity, convincing pharmaceutical companies to transition to more of an unbiased discovery approach (and maximizing their chance of novel discoveries) requires them to make a philosophical leap. That being said, we have had a lot of interest from groups within pharma and mature biotechs that are open to pursuing new ways to tackle disease, beyond drugging individual proteins.

LL: How do you think that your approach may provide something to the industry that target-based screening and synthetic libraries do not?

VC: On an immediate level, our platform unlocks millions of unique molecules for drug discovery. For example, our library contains > 5 million small molecules that you cannot access anywhere else – and we are just getting started. Given that natural products are more drug-like, have greater structural diversity, and are predicted to cover bioactive space more efficiently different than synthetic libraries, we effectively have a “drug discovery treasure chest” at our disposal. At a high level, we believe that many of these molecules represent inherently more translatable medicines – those that will have much more success in clinical trials.

The traditional drug discovery process has become very inefficient, with only one in 11 drug candidates passing clinical trials. Nearly all such failures are due to a lack of efficacy at an acceptable toxicity profile. By prioritizing our unique, more drug-like molecules with (a) human “priors” or hints of safe therapeutic activity from the use of source medicinal plants and (b) global-patterns of how millions of these molecules relate to different (beneficial and harmful) biological mechanisms, we are systematizing our chance at getting more drugs to work in people.

LL: Can you elaborate on Enveda’s next steps?

VC: We have a powerful drug discovery platform that can allow us to enable the translation of medicinal plant molecules into breakthrough medicines across a range of disease areas. We want to continue building our early-stage portfolio of candidates and soon build out a clinical development team around a focused therapeutic area to advance new drugs into clinical development. Outside of the target therapeutic area, we plan to continue finding partners for clinical development and beyond.

Viswa Colluru was speaking with Laura Elizabeth Lansdowne, Managing Editor for Technology Networks.

Viswa founded Enveda after leading cross-functional teams across innovation, product strategy, and translational biology as an early employee at Recursion Pharmaceuticals. He holds a Ph.D. in immuno-oncology from the University of Wisconsin-Madison, where he developed new therapeutic agents for the treatment of prostate cancer.